The purpose of this proposal is to examine whether women who have been exposed to chlorinated pesticides, which have been identified as endocrine disruptors in animals, have alterations in their ovarian function as measured by frequency of menstrual cycle disturbances. Such disturbances may impact the frequency of infertility and sub-fertility, as well as that of other hormone-dependent illnesses such as breast cancer and osteoporosis, which affect large numbers of women. The target exposed population is a group of Laotian immigrants who are frequent consumers of fish caught locally in the San Francisco Bay. Contamination levels above health protective advisories have been documented in Bay fish and shellfish. Immigration from Asia, as well as consumption of contaminated fish, have been shown to be associated with higher organochlorine pesticide body burdens. Exposure will be measured by determining levels of DDT and other pesticide metabolites in a serum sample. Menstrual function will be assessed by measuring urinary metabolites of steroid hormones daily during three menstrual cycles of approximately 50 Laotian women. We are currently studying menstrual function in a group of 400 women, so we have developed algorithms to determine ovulatory status, day of ovulation and menstrual cycle disturbances such as short luteal phase, which will also be applied to the new sample. This previously-collected sample will serve as a baseline comparison group, as well as a source of data on the effects of lifestyle factors (such as smoking and alcohol consumption) on hormone function. The availability of this previously assembled cohort allows for more efficient examination of a small group of exposed women. This will be one of the first studies to examine biomarkers of exposure in relation to patterns of the menstrual cycle assessed directly by daily hormone levels. In addition to serum pesticide levels, questionnaire measures of exposure (e.g. species-specific fish consumption) will be examined in relation to ovarian function defined as continuous measures (e.g. cycle length) and dichotomous "abnormalities" (e.g. short luteal phase or anovulation). If interesting associations are revealed, follow-up in larger populations, with a larger gradient of exposure, will be warranted.